PROJECT SUMMARY Individuals with intractable focal epilepsy and normal structural imaging (non-lesional focal epilepsy, NLFE) pose treatment challenges. When no lesion is detected, localization of the seizure focus is difficult and surgical success elusive. It is unknown how focal seizures arise in cortex with radiographically normal structure, and whether pathology below the resolution of standard 3T MRI is a common culprit. We recently demonstrated that somatic mutations in SLC35A2, which encodes a UDP galactose transporter, account for some NLFE. Some radiographically normal cases exhibited FCD1a, demonstrating that somatic SLC35A2 mutations can account for seizure onset in structurally normal and abnormal cortex, and identifying SLC35A2 as the first gene underlying FCD1a and NLFE. We recently found somatic brain variants in PLXNB1 and BRAF in cases with normal imaging but FCD2a on pathology. PLXNB1 may represent the first non-mTOR gene in FCD2. We thereby demonstrated brain somatic variants can result in radiographic NLFE, some with abnormal pathology. As gene identification illuminates the biology of focal epilepsy and informs therapy, discoveries can have major clinical implications. We will further explore the somatic genetics of radiographic NLFE, including 1) replication of gene discoveries in new cases, additional gene identification, and deeper sequencing for lower-frequency pathogenic alleles; 2) clarifying phenotypes associated with specific somatic mutations, 3) development of imaging to detect subtle abnormalities; 4) correlation of variant allele frequency (VAF) with pathology and EEG from specific biopsy sites within resected brain and 5) association of regional pathology with EEG when mutations are absent. In Aim 1 We will define the somatic genetic landscape of NLFE. In surgical epilepsy samples, we will conduct ultra-high-depth whole exome sequencing of brain resected from the seizure focus from individuals with NLFE to identify additional somatic mutations in SLC35A2, PLXNB1, BRAF, and other genes. In Aim 2 we will define genotype-phenotype correlations in NLFE. We will determine phenotypes associated with somatic mutations in specific genes in a cohort of individuals with NLFE, focusing on age at onset of epilepsy, abnormalities on presurgical advanced 3T and 7T MRI with computational post-processing, and histopathologic analysis of resected tissue. In Aim 3 we will define the regional EEG, pathological, and allelic burden of pathogenic somatic variants within NLFE. During epilepsy surgery, we will perform 4-6 biopsies from tissue destined for resection using an MRI-localized, electrophysiology-guided procedure. We will characterize pathology and EEG firing pattern in each biopsy, distinguishing the EEG-designated seizure focus (core) vs. surround (penumbra). We will quantitatively genotype each biopsy for variants deemed pathogenic to establish per-biopsy VAF and when mutations are absent, correlate EEG pattern with histopathologic abnormalities. This will define associations among VAF, EEG firing pattern, and pathology.